Barbara M. A.
van Vugt-Lussenburg
*a,
Daan S.
van Es
*b,
Matthijs
Naderman
a,
Jerome
le Notre
b,
Frits van der
Klis
b,
Abraham
Brouwer
ac and
Bart
van der Burg
b
aBioDetection Systems bv, Science Park 406, 1098XH Amsterdam, The Netherlands
bWageningen Food and Biobased Research (WFBR), Bornse Weilanden 9, 6708WG, Wageningen, The Netherlands. E-mail: daan.vanes@wur.nl
cVU University, Faculty of Sciences, Department of Animal Ecology, De Boelelaan 1085, 1081HV Amsterdam, The Netherlands
First published on 27th February 2020
FDCA esters are highly relevant biobased alternatives for currently used benzene dicarboxylic acid esters. Despite all the developments on 2,5-FDCA applications, to the best of our knowledge thus far no toxicological data were available for 2,5-FDCA esters. In the present study we aimed to fill this gap, by using an in vitro reporter gene assay approach to compare the activity profile of commonly used phthalates to that of their furan-based counterparts. The assay selection was aimed at the detection of endocrine activity, since several phthalates are heavily scrutinised for their endocrine disrupting properties. However, to avoid missing other relevant toxicological endpoints, several assays able to detect various forms of cellular stress were also included in the panel. The results showed that the (ortho)benzene dicarboxylic acid esters were predominantly active on several of the endocrine assays. In comparison, six of the seven furan dicarboxylic acid based diesters tested here showed no activity in any of the 13 assays used. Only the isobutyl derivative DIBF showed moderate estrogenic activity on one assay, compared to much more pronounced activities on four assays for the ortho-phthalate analogue. Overall, the results presented in this paper are a strong indication that 2,5-FDCA based diesters in general are not only technically viable alternatives to phthalates, but also offer significant toxicological benefits, which supports a non-regrettable substitution.
Currently, the dominant industrial polyester is polyethylene terephthalate (PET), well known for its application in e.g. beverage bottles, textile fibres, etc. While the diol component ethylene glycol is already produced from renewable biomass (on small industrial scale),4 the diacid component, i.e. biobased terephthalic acid (TA), is still under development.4–6 Given the enormous production volume of TA (>80 Mton per a),7,8 which is expected to continue to grow in the coming decades, biobased TA, or analogues, should be based on abundantly available biobased feedstocks, which do not compete with primary food production. Glucose, which can be obtained from polysaccharides like starch and cellulose, is therefore the most promising feedstock for the industrial production of biobased TA. Despite continuing efforts, the efficient conversion of highly functionalised sugars to TA remains a challenge.4 An alternative approach is the development of a functional analogue of TA, the biobased furan-2,5-dicarboxylic acid (or 2,5-FDCA).9 This diacid, which was already reported by Fittig in 187610 can be obtained from sugars with much higher efficiency than TA. While functionally comparable to TA, 2,5-FDCA shows different chemical and physical behaviour in various applications. When used in polyesters such as the PET analogue PEF (polyethylene-2,5-furanoate) it was for instance found that various properties, such as gas barrier properties, were significantly improved by using 2,5-FDCA as diacid.11–13 Many investigators have already shown that 2,5-FDCA based polyesters are functionally comparable to their TA based analogues, in applications such as films, foils, bottles and fibres.14,15 Other polymer applications of 2,5-FDCA include polyamides and unsaturated polyesters resins.15,16 As the dimethyl ester of terephthalic acid (DMT) is an important building block and hence a high-volume chemical substance, it is of particular interest to compare its toxicity profile with that of its (relatively) new FDCA based counterpart.
In analogy, also medium to long chain diesters of TA and phthalic acid (PA) are industrially important high volume chemicals, which mainly find use as plasticisers for PVC.17 Especially phthalate esters are subject to continuous scrutiny related to (potential) adverse health effects.18–20 However, complete substitution of disputed phthalates is challenging due to both economic as well as performance issues. Hence the development of cost effective, biobased, plasticisers that offer environmental, health and performance benefits compared to existing phthalates is highly desirable. 2,5-FDCA esters can potentially offer such benefits, as was already shown for their technical performance as plasticiser in PVC.21–23 As a next step, more insight is required into the potential health effects of these substances.
Given these developments, it can be expected that industrial 2,5-FDCA production will take off in the near future, and that the number of applications will grow, especially at first in applications where 2,5-FDCA has a clear performance benefit. This could include a positive ecological and toxicological profile. The replacement of disputed phthalate plasticisers by 2,5-FDCA esters is an interesting opportunity.
Despite all the developments on 2,5-FDCA applications, to the best of our knowledge thus far no toxicological data were available for 2,5-FDCA esters. An EFSA report on 2,5-FDCA, however, is available; the document reports 2,5-FDCA as ‘negative’ in in vitro genotoxicity tests, and a 90-day oral toxicity study in rats resulted in a NOAEL of 300 mg kg−1 day−1. Furthermore, no concern for accumulation in man was reported.24 In order to come to a non-regrettable long-term substitution of e.g. TA by 2,5-FDCA it is of utmost importance to get an early insight into the toxicity of these potentially high-volume chemicals.
This prompted us to study a broad series of FDCA esters (short to long chain) by means of in vitro human reporter gene assays. These in vitro human cell-based CALUX® reporter gene assays have been designed to detect interaction of a substance with a specific nuclear receptor or cell signalling pathway, rather than the overall effect of a substance on a complex biological system. The advantage of this approach is that the results are straightforward to interpret, since complicating factors such as metabolism, tissue distribution or receptor crosstalk do not play a role in these reporter gene assays. The panel generates an activity profile that can be used to provide clues on the possible mode-of-action of a substance, to visualise trends for structural analogues, and as a starting point for further investigation.25–32 The substances were analysed on a reporter gene assay panel covering a broad range of endpoints, including nuclear receptor hormone interaction, DNA damage, oxidative stress and cellular stress pathways. This panel of assays has been shown to be predictive for major human toxicological endpoints, including endocrine disruption, reproductive toxicity, genotoxicity and acute toxicity.25–32
Here we report on the in vitro toxicological effects of geometry and substitution of a range of isomeric benzene- and furan dicarboxylic acids and esters. In vitro analysis of the substances was performed on a panel of effect-based CALUX reporter gene assays.
In addition to these key assays directed specifically to the detection of endocrine active substances, several more general assays to assess toxicity of substances were included. The aryl hydrocarbon receptor (AhR) CALUX,58 for example, assesses toxicity of dioxin-like substances and PAHs. AP-1 CALUX detects substances which interfere with cell cycle control, while ESRE CALUX measures an early stage of unfolded protein response as a result of cellular stress. Activity on the Nrf2 CALUX is indicative for oxidative stress; several phenolic substances have been known to undergo redox cycling, which would trigger this particular assay.59 Finally, activation of p53 GENTOX CALUX is indicative for DNA damage.29,32
This resulted in a panel of thirteen CALUX assays (Table 1). To be confident that this selection would enable the detection all effects elicited by the benzene dicarboxylic acids without missing other important endpoints, these substances were also analysed on the twelve non-selected CALUX assays available in our research facility (anti-ERα, AR, GR, PR, TRβ, RAR, LXR, PAH, Hif1α, TCF, NFκB, p21). ESI Table 2† shows that none of these assays were activated by any of the test substances, which confirms that the current selection of assays was adequate for the purpose of this study.
Entry number | CAS | Substance | Cytotox | ERα | AR-anti | PR-anti | GR-anti | TRβ-anti | PPARα | PPARγ | AhR | AP-1 | ESRE | Nrf2 | p53 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Reference substance | −6.6 | −12.2 | −7.7 | −11.0 | −9.5 | −6.9 | −9.7 | −7.7 | −12.3 | −9.5 | −7.5 | −5.4 | −9.0 | |
2 | 620-92-8 | Bisphenol F | — | −6.6 | −5.4 | −4.8 | — | NA | — | — | — | — | −4.7 | — | −3.3 |
3 | 80-05-7 | Bisphenol A | −4.0 | −7.7 | −6.8 | −5.5 | −4.5 | −4.2 | — | — | — | — | −4.3 | — | NA |
4 | 140-66-9 | 4-tert-Octylphenol | −5.5 | −7.2 | −6.0 | −6.1 | — | — | — | — | -— | — | — | −4.5 | — |
5 | 104-40-5 | 4-Nonylphenol | −4.9 | −5.1 | −5.8 | −5.5 | — | — | — | — | — | — | −4.6 | — | NA |
6 | 4376-18-5 | Monomethyl phthalate | — | — | — | — | — | NA | — | — | — | — | — | — | NA |
7 | 131-70-4 | Monobutyl phthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
8 | 2528-16-7 | Monobenzyl phthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
9 | 4376-20-9 | Mono(2-ethylhexyl) phtalate | −3.5 | — | — | — | — | — | −5.5 | −4.7 | — | — | — | — | −3.2 |
10 | 131-11-3 | Dimethyl phthalate | — | — | −4.7 | −3.6 | — | — | — | — | — | — | — | — | — |
11 | 1459-93-4 | Dimethyl isophthalate | — | −3.3 | −3.1 | — | — | — | — | — | — | — | — | — | — |
12 | 120-61-6 | Dimethyl terephthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
13 | 84-66-2 | Diethyl phthalate | −3.5 | −4.0 | −5.0 | −4.3 | — | — | — | — | — | — | — | — | — |
14 | 636-53-3 | Diethyl isophthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
15 | 636-09-9 | Diethyl terephthalate | — | — | −4.0 | — | — | — | — | — | — | — | — | — | — |
16 | 84-74-2 | Dibutyl phthalate | — | −5.2 | −4.7 | −4.5 | — | — | — | — | — | — | — | — | — |
17 | 84-69-5 | Diisobutyl phthalate | −4.5 | −5.3 | −5.0 | −5.0 | — | — | — | — | — | — | — | — | — |
18 | 84-75-3 | Di(n-hexyl) phthalate | −3.5 | −5.0 | −5.0 | −5.5 | −4.5 | NA | — | — | −4.0 | −4.2 | — | — | NA |
19 | 84-61-7 | Dicyclohexyl phthalate | −4.5 | −5.3 | — | −5.4 | −5.1 | NA | — | — | NA | NA | — | — | — |
20 | 85-68-7 | Butylbenzyl phthalate | −3.9 | −6.3 | −5.6 | −5.5 | — | — | — | −4.5 | −3.7 | — | — | — | — |
21 | 117-81-7 | Di(2-ethylhexyl) phthalate (synthesized) | — | −3.9 | — | — | — | — | — | — | — | — | — | — | — |
22 | 117-81-7 | Di(2-ethylhexyl) phthalate (commercial) | — | −4.0 | — | — | — | — | — | — | — | — | — | — | — |
23 | 137-89-3 | Di(2-ethylhexyl) isophthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
24 | 6422-86-2 | Di(2-ethylhexyl) terephthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
25 | 117-84-0 | Dioctyl phthalate | — | — | — | — | — | NA | — | — | — | — | — | — | — |
26 | 28553-12-0 | Diisononyl phthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
27 | 26761-40-0 | Diisodecyl phthalate | — | — | — | — | — | — | — | — | — | — | — | — | — |
For the phenolic substances in the list, most of the activity observed was, as expected, on the nuclear hormone receptor assays. The bisphenols (entries 2–3) and alkyl phenols (entries 4–5) were mainly active as estrogen receptor agonists, and androgen- and progesterone receptor antagonists, with a potency in the (sub-) micromolar range. This corresponds with their known activity as endocrine active substances.44,46 4-tert-Octylphenol additionally activated the Nrf2 CALUX, a pathway indicative of oxidative stress. Indeed, oxidative stress has been reported for this substance.59
It is known that in vivo, di-substituted phthalates are readily hydrolysed to their mono-esters.60 In rodent in vivo experiments, exposure to the monoesters results in similar adverse effects as exposure to the diesters.36,61 As a result, the mono-substituted phthalates are thought to be the endocrine active metabolites of the corresponding diesters, displaying endocrine disrupting activity.34,36,41,42 Therefore, in the current study, we determined the in vitro activity profile of four monoesters of (ortho)phthalic acid (entries 6–9) as well as 17 diesters (entries 10–27). Of the four monoesters tested only mono(2-ethylhexyl) phthalate (MEHP) was active (entry 9), yet not on the endocrine assays but on the peroxisome proliferator-assays PPARα and PPARγ, which is in agreement with previous reports.53–55,62 It has been observed before that the phthalate monoesters appear to be able to exert effects in vivo such as reproductive tract anomalies and decreased testosterone synthesis, which are indicative of an anti-androgenic mode of action,35,36,61 but that the phthalates fail to show anti-androgenic activity in in vitro androgen receptor interaction assays,38,42,63,64 suggesting an indirect, non-receptor mediated mode of action.
While the short-chain diesters dimethyl- and diethyl phthalate showed no or low estrogenic- and anti-androgenic activity, the medium chain diesters (dibutyl (entry 16), diisobutyl (entry 17), di(n-hexyl) (entry 18), dicyclohexyl (entry 19) and butylbenzyl (entry 20)) showed estrogenic, anti-androgenic and anti-progestagenic activity with LOECs in the micromolar range. Butylbenzyl phthalate displayed the highest potency for all three endocrine assays. Further increasing the chain length to di(2-ethylhexyl) phthalate (entry 21/22) resulted in a marked decrease in estrogenic potency (10–100 fold), and no detectable anti-androgenic or anti-progestagenic activity. The long chain phthalates dioctyl, diisononyl and diisodecyl phthalate (entries 25–27) showed no activity at all on any of the assays.
These results are in agreement with what is known about these substances;34,41,42,65–68 high molecular weight phthalates such as diisononyl- and diisodecyl phthalate are included in REACH as ‘not toxic for human health’.69,70 The lower molecular weight phthalates dibutyl-, diisobutyl-, butylbenzyl- and di(2-ethylhexyl) phthalate on the other hand, are classified in REACH as ‘very dangerous’, in Category 1B “substances regarded as toxic to reproduction”. The two shortest phthalates in this test, the methyl and ethyl diesters, do show some adverse effects in vivo, but not the reproductive tract developmental abnormalities that have been described for the C4 to C6 chain diesters.66,68 However, in concordance with our results, estrogen receptor agonism and androgen receptor antagonism has been observed in vitro for diethyl phthalate.40,42,66,68
The most prominent representative of the C2–C6 diesters is butylbenzyl phthalate (BBP); in the current study this substance is able to act as an agonist on the estrogen receptor and as an antagonist on the androgen- and progesterone receptor with higher potency than any of the other phthalates tested. This correlates well with other studies; its endocrine disrupting activity has resulted in an industrial phase out of BBP.71,72
Although di(2-ethylhexyl) phthalate (DEHP) is highly scrutinised for its endocrine disrupting properties, two preparations of this substance (synthesized in-house (entry 21) versus commercially available (entry 22)) showed very little activity, apart from estrogen receptor agonism at relatively high LOEC (1 mM). Also the corresponding monoester (entry 9) and presumed active metabolite in vivo, did not show in vitro estrogen- or androgen receptor activation or antagonism. Although in vivo studies in rat, as well as epidemiological studies in men, suggest that DEHP/MEHP exposure results in reproductive tract anomalies and reduced sperm motility indicative for an anti-androgenic mode-of-action,38,61 other in vitro studies have also shown a lack of estrogen- and androgen receptor interaction.41,42,63,68 It has been suggested that these substances exert their effects via other mechanisms, for example by influencing steroidogenesis through CYP19 inhibition, rather than by direct interaction with the androgen receptor.61,68 This is supported by results of two in vitro assays, the H295R steroidogenesis assay and human testis explants, where both MEHP and DEHP were shown to significantly inhibit testosterone synthesis.61,73 For the isobutyl-, butyl- and butylbenzyl diesters, in vitro estrogen receptor transactivation has been reported previously, as well as in vivo endocrine effects,34,36,41,42,66,68 which is in line with our findings. Nonetheless, in the ECHA support document for BBP,74 its adverse effects are considered to be primarily related to effects on steroidogenesis, rather than through direct interaction with steroid hormone receptors.
The data in Table 1 clearly show a dependency of the endocrine activity of the ortho-phthalates on the ester chain length, which appears to have an optimum at C4-C6. Further increasing the chain length to C8 (DEHP entries 21/22 and DOP entry 25) results in a complete lack of activity. This could be due to the dramatic decrease in aqueous solubility of the longer chain phthalates (10−9 M for n-octyl versus 10−5M for n-butyl, see also ESI Table 4†), resulting in a significantly reduced bioavailability.75–78 Furthermore, Thomsen et al.79 reported an inverse relationship between phthalate solubility and temperature; e.g. for DBP solubility drops from 14.6 mg L−1 at 25 °C to 5.5 mg L−1 at 35 °C. Since our cell assays are performed at an incubation temperature of 37 °C, it is reasonable to assume that the actual solubilities for the phthalate esters are even lower than those reported for the 20–25 °C range. Hence, for the long chain phthalates (≥C8) (lack of) bioavailability could simply be the cause for observed lack of activity. A similar lack of bioactivity for long chain phthalates (≥C8) was reported by Ejlertsson et al. in the degradation of phthalic acid esters under methanogenic conditions.80
Another observation that can be made from the results in Table 1 is that there is no significant effect of ester chain branching on endocrine activity (compare DBP (entry 16) and DIBP (entry 17), or DHP (entry 18) and DCHP (entry 19), respectively). Note that in general chain branching leads to a slightly higher aqueous solubility; e.g. approximately two times higher for DIBP compared to DBP (see also ESI Table 4†).78
Overall it can be concluded that solubility, and hence bioavailability is probably the most important factor determining the bioactivity of the phthalate esters.
![]() | ||
Chart 1 Structures of benzene-, and related furan dicarboxylic acid (FDCA) isomers used in this study; TA = terephthalic acid, IPA = isophthalic acid, PA = phthalic acid. |
Substance | Anglea (°) | Db (Å) | pKa1c | pKa2c | Solubilityd | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
a Angle between carboxylic acid groups.82,85 b Distance between carboxylic acid groups.82 c Data from ref. 83. d Solubility under ambient conditions in mg ml−1 (data from https://pubchem.ncbi.nlm.nih.gov/). e NR: Not reported. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TA | 180 | 5.73 | 3.51 | 4.82 | 0.015 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
2,4-FDCA | 150 | 5.08 | 2.63 | 3.77 | NRe | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
IPA | 120 | 4.40 | 3.46 | 4.46 | 0.12 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
2,5-FDCA | 129 | 4.83 | 2.60 | 3.55 | 1 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
PA | 60 | 2.67 | 2.98 | 5.28 | 0.7 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
3,4-FDCA | 83 | 3.37 | 2.55 | 7.23 | NRe |
Hence, FDCA's are in general more polar than their benzene counterparts. Based on these differences, we expect different effects in the interactions with human nuclear receptors and cell signalling pathways.
Most of the diesters that are the focus of this study are prepared from the corresponding diacids. Furthermore, under environmental and in vivo conditions it can be expected that all diesters will (eventually) undergo hydrolysis to the monoesters first, and subsequently to the diacids.78,81,82 Hence, in order to exclude any toxicological effects of the parent diacids themselves, they were analysed on the CALUX assay panel.
Interestingly, none of the free acids used in this study showed any effect on the assays used (data not shown, see ESI Table 3†). This could imply that the free acids are not active in the cells. However, lack of activity could also be due to inability to enter the cells. Under physiological conditions, i.e. pH 7.2, all of the diacids used will be deprotonated given their pKa values (see Table 2), which could inhibit uptake in the cells, explaining the lack of observed effect.83,84
As discussed previously, while the dimethyl esters of phthalic acid (DMP) and isophthalic acid (DMIP) showed activity, no effects were found for the terephthalic acid (DMT) isomer. In contrast, none of the furan derivatives tested were active on any of the assays (Table 3, entries 28–33). The 2,5-FDCA and 3,4-FDCA dimethyl esters were commercial samples that were subsequently purified, while the 2,4-FDCA derivative was prepared and purified in our labs according to a previously published procedure (see ESI†).86
Entry number | CAS | Substance | Abbreviation | Cytotox | ERα | AR-anti | PR-anti | GR-anti | TRβ-anti | PPARα | PPARγ | AhR | AP-1 | ESRE | Nrf2 | p53 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
28 | 131-11-3 | Dimethyl phthtalate | DMP | — | — | −4.7 | −3.6 | — | — | — | — | — | — | — | — | — |
29 | 1459-93-4 | Dimethyl isophthalate | DMIP | — | −3.3 | −3.1 | — | — | — | — | — | — | — | — | — | — |
30 | 120-61-6 | Dimethyl terephthalate | DMT | — | — | — | — | — | — | — | — | — | — | — | — | — |
31 | 4282-33-1 | Dimethyl-3,4-furandicarboxylate | DM-3,4-FDCA | — | — | — | — | — | — | — | — | — | — | — | — | — |
32 | 4282-32-0 | Dimethyl-2,5-furandicarboxylate | DM-2,5-FDCA | — | — | — | — | — | — | — | — | — | — | — | — | — |
33 | 1710-13-0 | Dimethyl-2,4-furandicarboxylate | DM-2,4-FDCA | — | — | — | — | — | — | — | — | — | — | — | — | — |
34 | 84-66-2 | Diethyl phthalate | DEP | −3.5 | −4.0 | −5.0 | −4.3 | — | — | — | — | — | — | — | — | — |
35 | 636-53-3 | Diethyl isophthalate | DEIP | — | — | — | — | — | — | — | — | — | — | — | — | — |
36 | 636-09-9 | Diethyl terephthalate | DET | — | — | -4.0 | — | — | — | — | — | — | — | — | — | — |
37 | 53662-83-2 | Diethyl-2,5-furandicarboxylate | DEF | — | — | — | — | — | — | — | — | — | — | — | — | — |
38 | 84-69-5 | Diisobutyl phthalate | DIBP | −4.5 | −5.3 | −5.0 | −5.0 | — | — | — | — | — | — | — | — | — |
39 | N/A | Diisobutyl-2,5-furandicarboxylate | DIBF | — | −4.3 | — | — | — | — | — | — | — | — | — | — | — |
40 | 117-81-7 | Di(2-ethylhexyl) phthalate | DEHP (synthesized) | − | −3.9 | — | — | — | — | — | — | — | — | — | — | — |
41 | 117-81-7 | Di(2-ethylhexyl) phthalate | DEHP (commercial) | — | -4.0 | — | — | — | — | — | — | — | — | — | — | — |
42 | 137-89-3 | Di(2-ethylhexyl) isophthalate | DEHIP | — | — | — | — | — | — | — | — | — | — | — | — | — |
43 | 6422-86-2 | Di(2-ethylhexyl) terephthalate | DEHT | — | — | — | — | — | — | — | — | — | — | — | — | — |
44 | 158099-01-5 | Di(2-ethylhexyl)-2,5-furandicarboxylate | DEHF (method 1) | — | — | — | — | — | — | — | — | — | — | — | — | — |
45 | 158099-01-5 | Di(2-ethylhexyl)-2,5-furandicarboxylate | DEHF (method 2) | — | — | — | — | — | — | — | — | — | — | — | — | — |
46 | 26761-40-0 | Diisodecyl phthalate | DIDP | — | — | — | — | — | — | — | — | — | — | — | — | — |
47 | N/A | Diisodecylfuran-2,5-dicarboxylate | DIDF | — | — | — | — | — | — | — | — | — | — | — | — | — |
For the furan dicarboxylic acid esters no effect of isomerism, and hence of symmetry or dipole moment, is observed under the conditions tested in our assays.
A possible reason why the FDCA methyl esters show no effect while their benzene analogues do, could lie in a different susceptibility to (enzymatic) hydrolysis of these esters. Oae et al. reported significant differences between the rate of hydrolysis of dimethyl 2,5-FDCA and 3,4-FDCA versus DMT (17 and 1.3 times higher respectively), which could result in a relatively fast hydrolysis of the furan derivatives, successively to the respective mono esters and diacids.87 The latter were found to be completely inactive (vide supra).
In order to investigate the effect of ester chain length, next the diethyl esters of the phthalate family were compared with diethyl 2,5-FDCA (purified commercial material). As discussed in the previous chapter, diethyl phthalate and diethyl terephthalate showed moderate to low activity on the endocrine assays, while the isophthalic acid analogue was inactive. Once again, no activity was found for the analogous 2,5-FDCA derivative (Table 3, entries 34–37).
In the current study, we observed a significant increase in both the potency and the number of active assays for the C4 phthalate derivatives (Table 1, entries 16–17). To investigate if 2,5-FDCA diesters with similar chain length likewise show an increased activity compared to their diethyl- and dimethyl counterparts, the CALUX profile of diisobutyl phthalate was compared to the profile of diisobutyl-2,5-FDCA (Chart 2). While diisobutyl phthalate (DIBP) was the second most active phthalate derivative tested, showing micromolar-range activity as an estrogen, anti-androgen and anti-progestin, in contrast diisobutyl-2,5-FDCA (DIBF) acted as a weak estrogen only (Table 3, entries 38–39). While DIBF is the only active furan derivative tested thus far, the observed effects are not as pronounced as those observed with DIBP.
![]() | ||
Chart 2 Chemical structures of isobutyl and 2-ethylhexyl diester of PA (DIBP and DEHP) and 2,5-FDCA (DIBF and DEHF). |
Further increasing the alcohol chain length from isobutyl to 2-ethylhexyl (2EH), i.e. an extension of the isobutyl motif (Chart 2), results in a significant reduction of effect in our assays, as is apparent from the weak estrogenic activity only in the case of DEHP, with very similar LOEC values for the synthesized and the commercial preparation (Table 3, entries 40 and 41). The absence of any response of the iso- and terephthalate isomers in our assay (as opposed to the ortho-phthalate) shows that positional isomerism in the phthalate family has a significant impact (Table 3, entries 40–43). The furan-based analogue of DEHP, di(2-ethylhexyl)-2,5-FDCA, prepared for this study using two different methods, did not activate any of the CALUX assays (entries 44 and 45).
Further increasing the chain length of the ester group to isodecyl results in complete absence of effects, both for the phthalate (DIDP) and the FDCA diester (DIDF) (Table 3, entries 46 and 47). This observation is in line with the industrial move from DEHP as general purpose plasticiser to the longer chain analogues DINP and DIDP.70,88
All CALUX assays were performed, as described in detail in the publicly available DB ALM protocol 197, in assay medium, consisting of DMEM without phenol-red indicator (Gibco) supplemented with 5% DCC-stripped fetal calf serum, 1× non-essential amino acids (Gibco) and 10 U ml−1 penicillin and 10 μg ml−1 streptomycin. For seeding, a cell suspension in assay medium was made of 1 × 105 cells per ml (U2-OS) or 4 × 105 cells per ml (H4IIE), and the white 384-wells plates were seeded with 30 μl per well cell suspension using a MicroFlo Select dispenser (BioTek). After 24 h, exposure medium was prepared by adding 0.2% of test substance dilution series in DMSO to a 96-wells plate with assay medium. Of this exposure mixture, 30 μl was added to the assay plates containing the CALUX cells, resulting in final DMSO concentrations of 0.1%. The final concentrations of the substances in the wells were 1E−4–3E−10 M in 0.5 log unit increments. Additionally, DMSO blanks and a full dose response curve of the relevant reference substance were included on each plate. All samples were tested in triplicate. The preparation of the substance dilution series as well as the exposure of the cells were performed on a Hamilton Starlet liquid handling robot coupled to a Cytomat incubator. In order to be able to detect receptor antagonism, the assays were also performed in antagonistic mode. The assay procedure was as described above, with the only exception that the EC50 concentration of the reference agonist was present during the exposure.
After 24 h exposure the exposure medium was removed and 10 μl per well Triton-lysis buffer was added by the MicroFlo Select. Subsequently, the luciferase signal was measured in a luminometer (Infinite Pro reader coupled to a Connect stacker, both TECAN), essentially as described before.90
In the present study, an in vitro reporter gene assay approach was used to compare the activity profile of commonly used phthalates to that of their furan-based counterparts on a broad series of toxicological endpoints. The assay selection was aimed at the detection of endocrine activity, since several phthalates are heavily scrutinised for their endocrine disrupting properties. However, to avoid missing other relevant toxicological endpoints, several assays able to detect various forms of cellular stress were also included in the panel.
The results showed that the benzene dicarboxylic acid esters were predominantly active on the endocrine assays ERα, anti-AR and anti-PR, while effects on other endpoints such as peroxisome proliferation (PPARs) or genotoxicity (p53) were seen only sporadically. The results obtained for the benzene dicarboxylic acid esters correspond well with literature, showing that endocrine activity (mainly estrogenic and anti-androgenic) increases with ester chain length, reaching a maximum at C4-C6, while longer chains result in a rapid decrease in activity.
Furthermore, it was shown that positional isomerism in the phthalate family has significant effects, with ortho-phthalates being by far the most active substances.
In comparison, six of the seven furan dicarboxylic acid based diesters tested here showed no activity in any of the 13 assays used. Only the isobutyl derivative DIBF showed moderate estrogenic activity on one assay, compared to much more pronounced activities on four assays for the ortho-phthalate analogue.
As a follow-up of the current study it would be relevant to also assess whether the 2,5-FDCA based diesters are able to interfere with steroidogenesis, since it has been suggested for phthalates that they exert their adverse activity not only via direct interaction with nuclear steroid hormone receptors, but also by affecting steroidogenesis.61,68,73 This could be assessed using a steroidogenesis assay (H295R, OECD TG456), or with enzyme inhibition assays for the main enzymes involved in steroidogenesis, CYP17 and CYP19. Additionally, since plastics often end up in the environment, it would be advisable to investigate the biodegradation- and ecotoxicological properties of 2,5-FDCA based diesters and derived products. For example, Jia et al. have shown that PEDF-30 + co-polyesters are enzymatically degradable.92
Overall, the results presented in this paper are a strong indication that 2,5-FDCA based diesters in general are not only technically viable alternatives to phthalates, but also offer significant toxicological benefits, which supports a non-regrettable substitution.
Footnote |
† Electronic supplementary information (ESI) available: Full substance list, synthesis methods and full bioassay results tables. See DOI: 10.1039/c9gc04348a |
This journal is © The Royal Society of Chemistry 2020 |